VOL. 10, NO. 3, FEBRUARY 2015 ISSN 1819-6608
ARPN Journal of Engineering and Applied Sciences
© 2006-2015 Asian Research Publishing Network (ARPN). All rights reserved.
www.arpnjournals.com
1352
IMMUNE NETWORK ALGORITHM IN MONTHLY STREAMFLOW
PREDICTION AT JOHOR RIVER
Nur Izzah Mat Ali
1
, M. A. Malek
2
, Amelia Ritahani Ismail
3
1
Department of Civil Engineering, Universiti Tenaga Nasional, Kajang, Malaysia
2
Institute of Energy, Policy and Research (IEPRe), Universiti Tenaga Nasional, Kajang, Malaysia
3
Department of Computer Science, Kuliyyah of Information and Communication Technology, International Islamic University Malaysia,
Kuala Lumpur, Malaysia
E-Mail: izzah5255@gmail.com
ABSTRACT
This study proposes an alternative method in generating future stream flow data with single-point river stage.
Prediction of stream flow data is important in water resources engineering for planning and design purposes in order to
estimate long term forecasting. This paper utilizes Artificial Immune System (AIS) in modelling the stream flow of one
stations of Johor River. AIS has the abilities of self-organizing, memory, recognition, adaptive and ability of learning
inspired from the immune system. Immune Network Algorithm is part of the three main algorithm in AIS. The model of
Immune Network Algorithm used in this study is aiNet. The training process in aiNet is partly inspired by clonal selection
principle and the other part uses antibody interactions for removing redundancy and finding data patterns. Like any other
traditional statistical and stochastic techniques, results from this study, exhibit that, Immune Network Algorithm is capable
of producing future stream flow data at monthly duration with various advantages.
Keywords: immune network algorithm, artificial immune system, streamflow prediction.
INTRODUCTION
Streamflow forecasts is crucial to flood
mitigation and water assets administration and
arrangement. While transient expectation, for example,
hour or every day guaging is essential for surge cautioning
and resistance, long haul forecast focused around month to
month, occasionally or yearly time scales is exceptionally
helpful in store operations and watering system
administration choices, for example, planning discharges,
apportioning water to downstream clients, dry season
moderation and overseeing stream bargains or executing
conservative compliance [1].
A critical number of gauging models and
approaches have been created and connected with this
field because of the imperatives of hydrologic forecasting.
These streamflow forecasting models can be categorized
as process-driven methods and data-driven methods [2].
Linear models such as AutoRegressive (AR),
AutoRegressive Moving Average (ARMA),
AutoRegressive Integrated Moving Average (ARIMA),
and Seasonal ARIMA (SARIMA) had made a great
success in streamflow prediction[3]. Artificial Neural
Networks (ANNs), Genetic Algorithms and Artificial
Immune Systems (AIS) are some of streamflow prediction
techniques which have grown popularity lately.
In this study the anticipated future streamflow
information will be utilized for estimating of water assets
arranging and operational frameworks. This anticipated
streamflow information is extremely valuable for long
haul guaging in the arranging and operation of the water
assets administration. As an expansion, the utilization of
the propose technique i.e Artificial Immune System will
be another commitment to the field of hydrology in
anticipating month to month streamflow information.
The objective of this study is to develop and test the
feasibility and accuracy of the monthly streamflow
prediction model using an Artificial Immune System
(AIS).
ARTIFICIAL IMMUNE SYSTEM
AIS was characterized as versatile frameworks,
propelled by hypothetical immunology and watched
resistant capacities, standards and models, which are
connected to problem solving [4]. There are many
conceptions and opinion that have been taken out from the
biological immune systems to develop new set of
computer instructions to apply as authentic world
engineering and scientific quandaries solver.
The external microorganism is defended by the
immune system from attacking the human bodies, as it is
the main role of the immune system. [4]. Two types of
immune system immunity, which is innate and adaptive
immune system. Both systems are formed of two main
lines of defense in the immune system[5]. It is capable to
nearly recognize any pathogen or foreign or molecules and
eliminate them from body [6]. The main applications of
AIS that had been done before are data mining [7], pattern
recognition [8], anomaly detection [9] and scheduling
[10]. AIS has three main algorithms which are clonal
selection algorithm (CSA), immune network algorithm
(INA) and negative selection algorithm (NSA)
[11][12][13].
INA for the most part connected to manage
dynamic circumstance and improvement emergency where
NSA generally fruitful applying its methodologies in
abnormality identification [14]. Clonal Selection Principle
is satisfactory in taking care of the issue with respect to
scheduling and optimization [14]. This study uses INA in
AIS to foresee month to month streamflow data[1].